This paper examines purpose and rationale clauses, two subtypes of purposive constructions. The study of these constructions has been mainly developed for English within formal syntax. A purpose clause is a VP-internal adjunct containing a gap bound to the matrix object(i.e. controllee), while a rationale clause is a VP-external adjunct lacking a gap bounded to the matrix object. A similar approach to controller-pivot relationships for purpose has been adopted in previous studies in Role and Reference Grammar. Based on cross-linguistic data, I argue that the lexical manifestation of the controlled element is a language-specific feature, i.e. it can be covert or overt. In some languages, the two lexical manifestations are possible, i.e. a construction-specific property.
Traditional “projectionist” accounts of transitivity project the argument structure of a clause from the head verb. Some studies within Construction Grammar have shown this does not account for cases in which syntactic frames override a verb’s inherent transitivity, arguing instead that transitivity is determined by the syntactic construction. Such examples typically come from English and related languages in which many or most verbs freely occur in transitive or intransitive frames without any overt derivational morphology. However, in languages such as Tepehua (Totonacan), verbs have rigidly specified transitivity, with no such overriding of argument structure. Role and Reference Grammar treats argument structure as a projection from the composite logical structure, accounting for clause structure in both types of languages.
This paper examines the ability of a mimetic verb in Japanese (e.g. burabura ‘manner of swinging’ + suru ‘do’ → burabura-suru) to occur in different morphosyntactic environments. Following Van Valin (2013), it argues that two seemingly contradictory standpoints, a constructionist’s view (Tsujimura 2005) and a projectionist’s view (Kageyama 2007), are actually complementary. While an account of intransitive mimetic verbs requires no postulation of constructional schemas, as these verbs show straightforward cases of linking in simple sentences, the paper utilizes two constructional schemas to cover notable characteristics of transitive mimetic verbs: one for mimetics compounded by -sase ‘cause’ and the other for the colloquial alternate of the adverbial mimetic-transitive verb combination.
The goal of this chapter is to provide a Role and Reference Grammar (RRG henceforth) analysis of cleft sentences in Persian with respect to the role of interaction over the domains of syntax, semantics and information structure. Clefting is functionally motivated by thematically marked expression of a single proposition via bi-partition syntax in the sense that a sequence of two clauses including a matrix and a cleft clause is employed in order to bring into focus an element that can be otherwise expressed as a non-focal element in the unmarked predicate-focus structure. This deviation from the unmarked predicate-focus structure feeds cleft sentences with a non-isomorphic/non-compositional feature that can be accounted for in terms of a constructional framework that encompasses the morphosyntactic, semantic and informational properties that a cleft construction has. RRG is equipped with a bi-lateral syntax-semantics representation that is enriched with discourse-pragmatic considerations which enables it to facilitate processing of linguistic properties of cleft constructions by means of constructional schemas. Non-compositionality in a cleft construction implies that the semantic composition of the whole cleft construction is not the “computable sum of the meaning of its parts” (Lambrecht 1994: 230). That is why a construction-based scrutiny will be necessary to explain how the mapping between semantics and information structure in clefting can be elucidated with respect to the syntactic juxtaposition of the constituents. It will be also discussed that there should be a necessity to distinguish between emphatic in in clefting and anaphoric in in extraposition in the Persian language with regard to the role of information structure that is distinctively formalized in the syntactic templates of the two constructions on the basis of the RRG constructional framework.
Starting from the idea of a “holistic approach” (Van Valin 1980) based on text interpretation and communication analysis, the chapter sketches a radical, i.e. back to the roots, remodelling of standard RRG (Van Valin & LaPolla 1997; Van Valin 2005, 2010). It will be shown that a bidirectional linking algorithm (syntax-to-semantics and semantics-to-syntax), no matter how useful it may be for computational implementation, is not an adequate model of human communication. As Van Valin (2006) himself recognizes, the semantic as well as the syntactic representation are already infiltrated by one another. Thus, RRRG will abandon the linking algorithms and instead advocate for three structural levels of different complexity that assumedly function simultaneously: lexical items, syntactic-semantic event templates and construction schemas. As in standard RRG, general rules and principles operate at all levels. In RRG, the most prominent of these principles is the Actor-Undergoer-Hierarchy which is based on actionsart-driven Logical Structures (LS). However, LS prove to be too coarse-grained to describe the different activity degrees material to argument realization. Therefore, a finer-grained Activity Hierarchy will be introduced. The functioning of this centrepiece of RRRG will be illustrated with verbs of emotion (Kailuweit 2005, 2007, 2012a) at the level of lexical items and with anticausative constructions (Kailuweit 2011b, 2012b) at the level of constructional schemas.
It is now accepted that constructions exist at all levels in grammar from clausal syntax to word level morphology and even within the lexicon itself where lexical items themselves may be viewed as constructions. Constructions may also encompass lexical, semantic and pragmatic information. The semantics and/or pragmatics are not predictable from the set of lexical items in the construction. There is now recognition that the RRG account of constructions is a significantly under-utilised resource (Nolan 2013, 2012a, b, 2011). As functional linguists, the important empirical questions are: (1) How does the theory understand a construction? (2) What does a construction contain? (3) How do the constructions relate to the grammar and other constructions? (4) To what extent is our grammatical knowledge organized in constructions? (5) Do constructions include information about form, function and meaning? (6) Are constructions organized in a structured network? This paper proposes a view of constructions as structured grammatical objects and we motivate this account with new evidence of constructions from Modern Irish.
Framed within Role and Reference Grammar, this chapter provides a finer-grained account of the English resultative constructions (e.g. He hammered the metal flat/into a knife), which enriches the constructional schema suggested in Van Valin (2005). In so doing, we mainly follow the work carried out by Nolan (2011a, b) and Diedrichsen (2010, 2011), while also drawing on insights coming from the family of Construction Grammars (e.g. Goldberg 1995; Goldberg & Jackendoff 2004, inter alios). In turn, a further step is taken here by proposing the incorporation of additional features such as the motivation of the construction and its family resemblance connection; two essential issues which heavily depend upon the role of metaphor and metonymy (Ruiz de Mendoza & Mairal 2011).
Few researchers in natural language processing are nowadays concerned with linguistically-aware applications. On the contrary, the prevailing trend is towards the search of engineering solutions to practical problems, where researchers are motivated by the immediate gratification from the stochastic paradigm. As a result, there have been few attempts to confront the new challenges in linguistics from the natural language processing approach. The goal of this chapter is to introduce the theoretical foundation underlying ARTEMIS, a knowledge-based system which is intended to simulate natural language understanding in the framework of Role and Reference Grammar. More specifically, we will focus on how to enhance this functional model in order to make argumental constructions play a decisive role in the computational analysis of the deep semantics in the text.
This paper discusses how the Lexical Constructional Model (LCM) contributes to our understanding of meaning is constructed, interpreted and expressed. It especially addresses the role of constructional meaning in this complex process, while making critical revisions of other constructionist accounts of language, whether cognitivist or functionalist. It includes the notion of replicability into the definition of construction. According to this notion, a form-meaning pairing can be considered a construction, even if the pairing is not frequent, provided that it can be felt by competent native speakers as being ‘potentially replicable’, i.e. as being naturally meaningful without doing any violence to the nature of the language to which the construction belongs. The paper further argues that constructional structure mediates the syntactic realization of verbal meaning. In this view, meaning is not composed by assembling concepts, as postulated in Cognitive Grammar, but rather by making use of the conceptual scaffolding provided by constructions. Then, the paper relates the architecture of the LCM to a taxonomy of cognitive models and addresses meaning construction from the point of view of the descriptive and explanatory tools of the LCM. These tools include the definition of several central processes: subsumption, amalgamation, and saturation of variables. The role of each process is assessed at the various descriptive levels of the LCM. Finally, the paper relates formal expression to meaning representation in terms of idiomatic and non-idiomatic constructions. In this connection it specifies the requirements for full formal expression and relates such requirements to the format of constructional templates in the LCM.
This article discusses the concept of construction in the Lexical Constructional Model, focusing on tensions, concerned largely with the relationship between meaning and form in constructions, which have arisen in the model as a result of contributions from three different groups of scholars: functionalists, cognitivists and computational linguists/computer scientists. The article examines precursors of the LCM which made use of ideas from Role and Reference Grammar, and then assesses the influence of the cognitivist contribution and the later input from the computational knowledge base, FunGramKB, also based partly on RRG. Finally, it summarises recent work in RRG, suggesting that the new proposals for treating constructional schemas as central to RRG could prove useful in working towards solutions of the problems arising within the LCM.
The paper will explore the theoretical scope of the concept “construction”, as envisaged in Constructional approaches to grammar. Starting from the Role and Reference Grammar notion of Constructions, as represented in “Constructional Schemas”, it will be argued that Constructional Schemas as representations of linguistic knowledge can be used not only for language specific constructions, but for the wide range of argument structure and sentence structure constructions as well. This will be exemplified by extensive discussions of two well known German construction types, which are the bekommen-passive, a passive three-place argument structure construction, and the bracket structure, which is a sentence structure pattern that forms the basis of many syntactic phenomena in German. It will be argued that the Construction in this sense is to be treated as a “grammatical object”, whose use is systematically constrained by context factors and also by lexical-semantic factors. The Constructional Schemas give an extensive representation of the constructions by providing the constraints of their use, the constraints for their recognition in a stream of speech or writing, their syntax, their semantics, morphology and pragmatics. The model of constructional schemas caters for real-time processing in a workspace. The notion of constructional knowledge for the processing of linguistic utterances is then taken a step further and expanded to include constructions whose use and functionality is not mainly based on grammatical knowledge, but rather on cultural knowledge. Some “idioms” do not “work” on their own, but require a very subtle mix of culturally acquired background knowledge and situational factors, and their use is deeply embedded in basic behavioural patterns in a society of speakers. I will introduce three speech act constructions, which do not exhibit the form-function correlation that is generally described for the linguistic realisation of illocutionary force, and explain their pragmatic effects by adhering to Dawkin’s notion of the cultural unit “meme” and Wittgenstein’s idea of “life form”.
This paper examines purpose and rationale clauses, two subtypes of purposive constructions. The study of these constructions has been mainly developed for English within formal syntax. A purpose clause is a VP-internal adjunct containing a gap bound to the matrix object(i.e. controllee), while a rationale clause is a VP-external adjunct lacking a gap bounded to the matrix object. A similar approach to controller-pivot relationships for purpose has been adopted in previous studies in Role and Reference Grammar. Based on cross-linguistic data, I argue that the lexical manifestation of the controlled element is a language-specific feature, i.e. it can be covert or overt. In some languages, the two lexical manifestations are possible, i.e. a construction-specific property.
Traditional “projectionist” accounts of transitivity project the argument structure of a clause from the head verb. Some studies within Construction Grammar have shown this does not account for cases in which syntactic frames override a verb’s inherent transitivity, arguing instead that transitivity is determined by the syntactic construction. Such examples typically come from English and related languages in which many or most verbs freely occur in transitive or intransitive frames without any overt derivational morphology. However, in languages such as Tepehua (Totonacan), verbs have rigidly specified transitivity, with no such overriding of argument structure. Role and Reference Grammar treats argument structure as a projection from the composite logical structure, accounting for clause structure in both types of languages.
This paper examines the ability of a mimetic verb in Japanese (e.g. burabura ‘manner of swinging’ + suru ‘do’ → burabura-suru) to occur in different morphosyntactic environments. Following Van Valin (2013), it argues that two seemingly contradictory standpoints, a constructionist’s view (Tsujimura 2005) and a projectionist’s view (Kageyama 2007), are actually complementary. While an account of intransitive mimetic verbs requires no postulation of constructional schemas, as these verbs show straightforward cases of linking in simple sentences, the paper utilizes two constructional schemas to cover notable characteristics of transitive mimetic verbs: one for mimetics compounded by -sase ‘cause’ and the other for the colloquial alternate of the adverbial mimetic-transitive verb combination.
The goal of this chapter is to provide a Role and Reference Grammar (RRG henceforth) analysis of cleft sentences in Persian with respect to the role of interaction over the domains of syntax, semantics and information structure. Clefting is functionally motivated by thematically marked expression of a single proposition via bi-partition syntax in the sense that a sequence of two clauses including a matrix and a cleft clause is employed in order to bring into focus an element that can be otherwise expressed as a non-focal element in the unmarked predicate-focus structure. This deviation from the unmarked predicate-focus structure feeds cleft sentences with a non-isomorphic/non-compositional feature that can be accounted for in terms of a constructional framework that encompasses the morphosyntactic, semantic and informational properties that a cleft construction has. RRG is equipped with a bi-lateral syntax-semantics representation that is enriched with discourse-pragmatic considerations which enables it to facilitate processing of linguistic properties of cleft constructions by means of constructional schemas. Non-compositionality in a cleft construction implies that the semantic composition of the whole cleft construction is not the “computable sum of the meaning of its parts” (Lambrecht 1994: 230). That is why a construction-based scrutiny will be necessary to explain how the mapping between semantics and information structure in clefting can be elucidated with respect to the syntactic juxtaposition of the constituents. It will be also discussed that there should be a necessity to distinguish between emphatic in in clefting and anaphoric in in extraposition in the Persian language with regard to the role of information structure that is distinctively formalized in the syntactic templates of the two constructions on the basis of the RRG constructional framework.
Starting from the idea of a “holistic approach” (Van Valin 1980) based on text interpretation and communication analysis, the chapter sketches a radical, i.e. back to the roots, remodelling of standard RRG (Van Valin & LaPolla 1997; Van Valin 2005, 2010). It will be shown that a bidirectional linking algorithm (syntax-to-semantics and semantics-to-syntax), no matter how useful it may be for computational implementation, is not an adequate model of human communication. As Van Valin (2006) himself recognizes, the semantic as well as the syntactic representation are already infiltrated by one another. Thus, RRRG will abandon the linking algorithms and instead advocate for three structural levels of different complexity that assumedly function simultaneously: lexical items, syntactic-semantic event templates and construction schemas. As in standard RRG, general rules and principles operate at all levels. In RRG, the most prominent of these principles is the Actor-Undergoer-Hierarchy which is based on actionsart-driven Logical Structures (LS). However, LS prove to be too coarse-grained to describe the different activity degrees material to argument realization. Therefore, a finer-grained Activity Hierarchy will be introduced. The functioning of this centrepiece of RRRG will be illustrated with verbs of emotion (Kailuweit 2005, 2007, 2012a) at the level of lexical items and with anticausative constructions (Kailuweit 2011b, 2012b) at the level of constructional schemas.
It is now accepted that constructions exist at all levels in grammar from clausal syntax to word level morphology and even within the lexicon itself where lexical items themselves may be viewed as constructions. Constructions may also encompass lexical, semantic and pragmatic information. The semantics and/or pragmatics are not predictable from the set of lexical items in the construction. There is now recognition that the RRG account of constructions is a significantly under-utilised resource (Nolan 2013, 2012a, b, 2011). As functional linguists, the important empirical questions are: (1) How does the theory understand a construction? (2) What does a construction contain? (3) How do the constructions relate to the grammar and other constructions? (4) To what extent is our grammatical knowledge organized in constructions? (5) Do constructions include information about form, function and meaning? (6) Are constructions organized in a structured network? This paper proposes a view of constructions as structured grammatical objects and we motivate this account with new evidence of constructions from Modern Irish.
Framed within Role and Reference Grammar, this chapter provides a finer-grained account of the English resultative constructions (e.g. He hammered the metal flat/into a knife), which enriches the constructional schema suggested in Van Valin (2005). In so doing, we mainly follow the work carried out by Nolan (2011a, b) and Diedrichsen (2010, 2011), while also drawing on insights coming from the family of Construction Grammars (e.g. Goldberg 1995; Goldberg & Jackendoff 2004, inter alios). In turn, a further step is taken here by proposing the incorporation of additional features such as the motivation of the construction and its family resemblance connection; two essential issues which heavily depend upon the role of metaphor and metonymy (Ruiz de Mendoza & Mairal 2011).
Few researchers in natural language processing are nowadays concerned with linguistically-aware applications. On the contrary, the prevailing trend is towards the search of engineering solutions to practical problems, where researchers are motivated by the immediate gratification from the stochastic paradigm. As a result, there have been few attempts to confront the new challenges in linguistics from the natural language processing approach. The goal of this chapter is to introduce the theoretical foundation underlying ARTEMIS, a knowledge-based system which is intended to simulate natural language understanding in the framework of Role and Reference Grammar. More specifically, we will focus on how to enhance this functional model in order to make argumental constructions play a decisive role in the computational analysis of the deep semantics in the text.
This paper discusses how the Lexical Constructional Model (LCM) contributes to our understanding of meaning is constructed, interpreted and expressed. It especially addresses the role of constructional meaning in this complex process, while making critical revisions of other constructionist accounts of language, whether cognitivist or functionalist. It includes the notion of replicability into the definition of construction. According to this notion, a form-meaning pairing can be considered a construction, even if the pairing is not frequent, provided that it can be felt by competent native speakers as being ‘potentially replicable’, i.e. as being naturally meaningful without doing any violence to the nature of the language to which the construction belongs. The paper further argues that constructional structure mediates the syntactic realization of verbal meaning. In this view, meaning is not composed by assembling concepts, as postulated in Cognitive Grammar, but rather by making use of the conceptual scaffolding provided by constructions. Then, the paper relates the architecture of the LCM to a taxonomy of cognitive models and addresses meaning construction from the point of view of the descriptive and explanatory tools of the LCM. These tools include the definition of several central processes: subsumption, amalgamation, and saturation of variables. The role of each process is assessed at the various descriptive levels of the LCM. Finally, the paper relates formal expression to meaning representation in terms of idiomatic and non-idiomatic constructions. In this connection it specifies the requirements for full formal expression and relates such requirements to the format of constructional templates in the LCM.
This article discusses the concept of construction in the Lexical Constructional Model, focusing on tensions, concerned largely with the relationship between meaning and form in constructions, which have arisen in the model as a result of contributions from three different groups of scholars: functionalists, cognitivists and computational linguists/computer scientists. The article examines precursors of the LCM which made use of ideas from Role and Reference Grammar, and then assesses the influence of the cognitivist contribution and the later input from the computational knowledge base, FunGramKB, also based partly on RRG. Finally, it summarises recent work in RRG, suggesting that the new proposals for treating constructional schemas as central to RRG could prove useful in working towards solutions of the problems arising within the LCM.
The paper will explore the theoretical scope of the concept “construction”, as envisaged in Constructional approaches to grammar. Starting from the Role and Reference Grammar notion of Constructions, as represented in “Constructional Schemas”, it will be argued that Constructional Schemas as representations of linguistic knowledge can be used not only for language specific constructions, but for the wide range of argument structure and sentence structure constructions as well. This will be exemplified by extensive discussions of two well known German construction types, which are the bekommen-passive, a passive three-place argument structure construction, and the bracket structure, which is a sentence structure pattern that forms the basis of many syntactic phenomena in German. It will be argued that the Construction in this sense is to be treated as a “grammatical object”, whose use is systematically constrained by context factors and also by lexical-semantic factors. The Constructional Schemas give an extensive representation of the constructions by providing the constraints of their use, the constraints for their recognition in a stream of speech or writing, their syntax, their semantics, morphology and pragmatics. The model of constructional schemas caters for real-time processing in a workspace. The notion of constructional knowledge for the processing of linguistic utterances is then taken a step further and expanded to include constructions whose use and functionality is not mainly based on grammatical knowledge, but rather on cultural knowledge. Some “idioms” do not “work” on their own, but require a very subtle mix of culturally acquired background knowledge and situational factors, and their use is deeply embedded in basic behavioural patterns in a society of speakers. I will introduce three speech act constructions, which do not exhibit the form-function correlation that is generally described for the linguistic realisation of illocutionary force, and explain their pragmatic effects by adhering to Dawkin’s notion of the cultural unit “meme” and Wittgenstein’s idea of “life form”.